Sambar Deer (Cervus unicolor)

Range Description:

The Sambar extends from India and Sri Lanka east along the southern Himalayas (including Nepal and Bhutan) through much of south China (including Hainan Island) to Taiwan (where it occurs in the central and eastern parts; Lin, C.-Y. and Lee, L.-L. pers. comm. 2008). Further south it occurs in Bangladesh, throughout mainland South-east Asia (Myanmar, Thailand, Lao PDR, Cambodia, Vietnam, West Malaysia) and many of the main islands of the Greater Sundas (excepting Java): Sumatra, Siberut, Sipora, Pagi and Nias islands (all Indonesia), and Borneo (Malaysia, Indonesia, and Brunei) (Grubb 2005). The current distribution is now highly fragmented in much of this range (see Population). Payne et al. (1985) also listed the Philippines, but the Sambar does not occur there. A record from Ujung Kulon, Java, in van Schaik and Griffiths (1996: 107) is presumably an error for Javan Rusa R. timorensis. The Sambar has been introduced widely outside its native range, e.g.: San Luis Obispo Country, California; the Gulf Prairies and Edwards Plateau regions of Texas (Ables and Ramsey 1974); the St. Vincent Islands, Franklin Country, Florida (Lewis et al. 1990); Australia (Slee 1984; Freeland 1990); New Zealand (Kelton and Skipworth 1987); and Western Cape Province, South Africa (Lever 1985). These introduced populations are not included in the distribution map.
Countries: Native:
Bangladesh; Bhutan; Brunei Darussalam; Cambodia; China (Guangxi, Guizhou, Hainan, Hunan, Jiangxi, Sichuan, Yunnan); India; Indonesia (Sumatera); Lao People's Democratic Republic; Malaysia; Myanmar; Nepal; Sri Lanka; Taiwan, Province of China; Thailand; Vietnam
Australia; New Zealand; Saint Vincent and the Grenadines; South Africa; United States (California, Florida, Texas)


Almost everywhere outside Taiwan, the Sambar is in rapid decline leading to a widespread distribution of very low numbers and many local-level extinctions, with occasional healthy populations reflecting special circumstances. Few surveys specifically target Sambar, but a reasonable picture of its current status is emerging through surveys for Tiger Panthera tigris and through general wildlife surveys. Tiger-oriented camera-trapping predictably yields photographs of Sambar when they are present, and most Tiger surveys afford priority to elucidating Sambar status as a key prey, and therefore population determinant, of Tiger. Understanding Sambar status through surveys of Tigers or, sometimes, of Leopard P. pardus (e.g. Ngoprasert et al. 2007) needs care, because such surveys have generally selected the best remaining areas for very large mammals within the country or region in question, so as to maximize chances of picking up Tigers. They therefore cannot be seen as indicative of Sambar status in the wider landscape. Most importantly, in the several cases (identified below) where such surveys have found few Sambar, these indicate cause for alarm, because there is little reason to believe that there ought to be larger populations in many other areas in the country or region in question, even though large parts of it may remain unsurveyed. Occasional herds can survive in unexpected places through atypical site-specific circumstances in some countries (e.g. Tungittiplakorn and Dearden 2002). Populations outside the native range are not treated here, as they are irrelevant to assigning the Red List category and criteria. In Sri Lanka, numbers have probably declined substantially through poaching which increased during the civil war. Poaching occurred widely across the country, including within protected areas, particularly where there are concentrations of resettled refugees (Santiapillai and Wijeyamohan 2003). In India, although the Sambar occurs widely and in many habitat types, and large populations occur in well-secured protected areas, nowhere is it now regionally abundant (Sankar and Acharya 2004). It has been recorded in 208 protected areas (National Wildlife Database, Wildlife Institute of India, cited in Sankar and Acharya 2004); its distribution outside protected areas is now highly scattered. The reported ecological densities of Sambar in India mostly fall within 1–10 animals per km˛ within the protected area network, and depending on the varying levels of protection efficacy: Bhadra Tiger Reserve, 0.89 +/- 0.23 (SE) per km˛ (Jathanna et al. 2003); Madhya Pradesh Pench National Park dry deciduous forest, 9.6 animals per km˛ (Karanth and Nichols 2000); Kanha moist deciduous forest, 1.5 animals per km˛ (Karanth and Nichols 2000); Nagarahole moist deciduous forest, 5.5 animals per km˛ (Karanth and Sunquist 1992); Bandipur dry deciduous forest, 5.6 animals per km˛ (Karanth and Nichols 2000); Tadoba-Andhari dry deciduous forest, 3.3 animals per km˛ (Karanth and Kumar 2005); Melghat dry deciduous forest, 2.7 animals per km˛ (Karanth and Kumar 2005); Maharashtra Pench dry deciduous forest, 5.9 animals per km˛ (Karanth and Kumar 2005); Ranthambore semi-arid dry deciduous forest, 10.7 animals per km˛ (Kumar 2000); and Gir semi-arid dry deciduous forest, two animals per km˛ (Khan et al. 1996). Similar surveys at Kaziranga found too few Sambar to estimate populations there (Karanth and Nichols 2000), this presumably representing habitat characters rather than defective protection, given the buoyant populations of other deer at that site (Hog Deer Axis porcinus and Barasingha Rucervus duvaucelii). Outside protected areas, Sambar is present mostly in very low numbers, although larger numbers can still be found where its habitat is almost inaccessible to people. The recorded Bhadra density is low, reflecting poaching and forest-resource extraction (Jathanna et al. 2003), and the population density is steadily increasing following removal of these pressures in 2003 (K.U. Karanth and N.S. Kumar unpublished data). Nepal supports an important population in Bardia National Park (Dinerstein 1979). This remains healthy and there are also good populations in Parsa Wildlife Reserve, Chitwan National Park, and Sukla Phanta Wildlife Reserve. Across the country Sambar has declined so rapidly and so widely that it is now very rare outside protected areas, mostly recorded as isolated reports from forests adjoining these protected areas, and from south-east Nepal. Sambar has been victim of poachers, probably because of its large size (thus, more meat per animal and, for males, a more impressive trophy), perhaps compounded by nocturnal and relatively docile nature; recent declines are of the magnitude to meet IUCN criteria for Vulnerable on a national assessment (all information: Hem Sagar Baral pers. comm. 2008). It is unclear whether the demand for antler in velvet in eastern Asia is stimulating poaching in Nepal. No information concerning current status was traced from Bhutan but (simply on an area basis) numbers are unlikely to be large enough to influence the overall global population trend. In Bangladesh, Sambar was common in north-east, south-east and hill-tract forests in the 1960s (Khan 1985) but now only a very few persist in the south-east: “habitat loss and hunting pressure made them critically endangered in the country. The pressure became severe in the 1970s. I see little hope for this species in the next 30 years” (Md Anwarul Islam in litt. 2008). In China there were major declines in Sambar over much of the twentieth century, and populations probably decreased by more than 50% in the 30 years from 1978; records from Guangxi suggest that, there at least, the 1970s–1980s was the era of fastest decline (J. Fellowes pers. comm. 2008; B.P.L. Chan pers. comm. 2008). In some provinces populations seem now to have stabilized, and in some areas, particularly in Hainan (where several apparently stabilised populations are present, but all are small, reflecting habitat patch size: the largest may be of only a few hundred animals) and north Guangdong province (perhaps also the east of that province, where there is a nature reserve specifically for the species), populations may even have been increasing once again in the last few years. This is not so for Guizhou and the provinces between Guangdong and Yunnan (B.P.L. Chan and M.W.N. Lau pers. comm. 2008). Further north, in the provinces of Fujian, Jiangxi, Zhejiang, Hunan and Hubei, a multi-method Tiger survey in 2001 found that Sambar is very rare overall and scattered in present-day distribution (Tilson et al. 2004). For elsewhere in China, Smith and Xie Yan (2008) mapped wide presence in Sichuan, plus isolated populations in the provinces of Qinghai, Tibet, Chongqing, and Fujian. The original sources are not clear, and R. Harris (pers. comm.) doubts that Sambar has lived anywhere in Qinghai in recent times.  In Taiwan, numbers and geographic distribution reduced rapidly through hunting before 1987, then expanded somewhat since the mid 1990s reflecting reductions in hunting. The average population density in the Yu-Li Wildlife Refuge was 10.6 animals per km˛ in 2005. In some areas at high altitude, densities exceed 40 animals per km˛, and the animals have a large impact on the coniferous saplings in this area. But at low to mid altitude, hunting pressure remains high enough to restrain their distribution and density. Based on the known geographical distribution, there are five populations: A-Yu Mountain, Nanhu, Shei-Pa, Jade Mountains–Central Mountains, and Tawu. The A-Yu population was re-introduced (about 70 animals) by local deer farmers in 1988–1991, when the velvet price was very low through the impact of deer TB in 1985. This population has not expanded reflecting continuous poaching; it may be genetically polluted. The other four populations are all natural, assumed to be isolated by cross-island roads, and may have very low gene flows with each other. Sambar are largely restricted to protected areas, only occurring outside them in a few nearby areas. The population sizes and distributions in these areas are smaller than in the protected areas. The poaching pressure is an important limitation to Sambar in these unprotected areas (all information: Lin, C.-Y. and Lee, L.-L. pers. comm. 2008). In Myanmar, Sambar was detected at 12 of 15 camera-trapped Tiger sites during 1999–2002, indicating that it is still widespread (Lynam 2003). The Myanmar Forest Department and Wildlife Conservation Society have jointly now (as of May 2008) camera-trapped 21 survey areas (including the earlier 15) and found Sambar in 13 (Saw Htun pers. comm. 2008). This lower ratio of proven presence indicates how the areas selected for Tiger survey were above average for prey and may indicate increasing local-level extinctions. For example, in a multi-method study of one area not expected to hold Tiger, hunted animals showed that Sambar still occurred in Naungmung and Machanbaw (south of Hkakaborazi National Park, Kachin state), but it was so scarce that it was not detected by substantial amounts of either camera trapping or sign surveys (Rao et al. 2005). Even within these best areas, large declines are suspected and overall national population losses of about 50% over the last 25–30 years and in the following such period are likely (Than Zaw pers. comm. 2006; Saw Htun pers. comm. 2008).

In Thailand, several protected areas maintain big Sambar populations; these are mostly large with very few or no people living within them. Over much of the country Sambar is extinct or very nearly so, including in many protected areas, particularly in the north where most parks are small and heavily poached (Anak Patannavibool pers. comm. 2008). Sambar is rare in most areas of Thung Yai Naresuan, where commercial hunting has targeted large ungulates (Steinmetz et al. 2008). Sambar was recorded at over 40% of camera-trap locations during a Leopard survey in Kaeng Krachan National Park, and clearly remains widespread there. Abundance did not vary with proximity to villages, even though a variety of hunting methods occurs on site (Ngoprasert et al. 2007). Sambar is also locally abundant in Khao Yai National Park, Thap Lan National Park and Phu Khieo Wildlife Sanctuary (Lynam et al. 2006). In Huai Kha Khaeng Wildlife Sanctuary current Sambar density is about 2–3 animals per km˛ and it seems to be increasing (Anak Patannavibool pers. comm. 2008).

In Lao PDR, Sambar was described as very common throughout the wooded parts by Delacour (1940). Some decades later, the species was reported during 86% of 1988–1993 village interviews (n = 328) and was then still widespread throughout, except in the most heavily settled areas (Duckworth et al. 1999: Annex 5). However, sightings on surveys during 1992–1998 were very rare, and while calls heard by night and signs indicate a wide distribution, numbers are heavily depressed (Duckworth et al. 1999; Timmins and Ou 2001). Only in Nam Et–Phou Louey NPA did Sambar seem at least locally abundant (Davidson 1998), and reasonable populations were confirmed still to persist in 2003–2004 (Johnson et al. 2004); an exceptional area in an Indochinese context, that still even supports significant Tiger numbers. The Nakai Plateau and the mountainous interior of Nakai-Nam Theun NPA and a few other, localized, areas of Lao PDR, including the Dong Kalo tract of Xe Pian NPA, the Nam Hiang area of the Bolaven Plateau, and parts of Nam Kading NPA also seemed to support relatively high densities during the mid 1990s. The few sites resurveyed during 2004–2007 have in general shown major declines (e.g. Timmins and Robichaud 2005). There was not a single camera-trap photograph during fairly intensive camera-trapping in Nakai–Nam Theun NPA during 2006–2007, although the species has subsequently been found to persist in a few parts of the protected area (Johnson and Johnston 2007; W.G. Robichaud and Nam Theun Watershed Management Protection Authority pers. comm. 2008) and it is not clear whether quirks of camera-trap placement (e.g., hypothetically, a focus on slopes rather than valley bottoms) may have been a major contributor to the lack of photographs. The Nakai Plateau still supports noteworthy numbers in a Lao context, but even here there were signs of decline (Dersu 2007; R.J. Timmins pers. comm. 2008). Surprisingly, given the proximity to China and Viet Nam, a good population was found to persist in Phou Dendin NPA in the far north-east of Lao PDR in 2004–2005 (W.G. Robichaud pers. comm. 2008). In Vietnam, in most of the northern highlands, in particular, and the northern and central Annamites to an increasing extent, Sambar is very rare or has been hunted out from many areas of otherwise suitable habitat (Timmins et al. 1999; Grieser Johns 2000; Timmins and Trinh Viet Cuong 2001; R.J. Timmins pers. comm. 2008; B. Long pers. comm. 2008). Camera-trapping studies in Pu Mat National Park, Thua Thien Hue province and Quang Nam province between 1998 and 2007 recorded only a handful of records and local communities in all these areas report a major decline in numbers (B. Long pers. comm. 2008). With so many higher-profile species long listed as threatened on the Red List, many of them with restricted ranges, the status of Sambar in Viet Nam has largely been ignored, and the paucity of evidence during many surveys has rarely been highlighted in reviews. As with many large mammals, populations are thought to be best in southern parts of the country. It remains common in Cat Tien National Park (Polet and Ling 2004), but this is very much the exception, and it was not considered at all common in surveys of several areas of prime habitat in Dak Lak Province in 1997 (Le Xuan Canh et al. 1997). In Cambodia, Sambars “appear to be at least relatively common in some areas, though entirely absent from others, a feature which appears to have a strong correlation to local hunting pressure...At the current rate of persecution, this species is likely to decline sharply over the next few years” (Walston 2001). Sambar is still widespread, although heavily targeted by hunters and most populations are now very depressed (Timmins and Men 1999; Timmins et al. 2003; Tordoff et al. 2005: 171; Timmins 2006). In the eastern plain of Cambodia, where populations of many other ungulates are magnitudes more numerous than in other areas of Indochina, Sambar numbers are clearly low, almost certainly as a result of targeted hunting: numbers in some areas seem depressed even well below those of wild cattle (Timmins and Ou 2001). Protection activities in a few localised areas appear to have stabilised declines, but this is an exception to a general scenario of rapid decline. These declines are evidently faster and recent than in either Lao PDR or Viet Nam, reflecting the rapid pace of change of economic and logistical factors in Cambodia since the mid 1990s. In West Malaysia, Sambar was regularly camera-trapped in Taman Negara (Kawanishi and Sunquist 2004); population densities were taken to be 0.01–0.20 per km˛ (Kawanishi and Sunquist 2004). It is unclear how close this estimate is to a pre-exploitation density, because although these authors adjudged large mammal poaching to be negligible in Taman Negara, high levels were confirmed there by Wan Shahruddin (1998); whether Sambar was being targeted is unclear. Moreover, camera-trapping (suitable in methodology and intensity to find Sambar if any significant numbers were present) did not record Sambar in a fourth sector of Taman Negara in 1999 (Lynam et al. 2007). The great current rarity of Sambar in Peninsular Malaysia was evinced by a camera-trap survey for Tiger: it was photographed at only one of the nine areas surveyed during 1997–1999, Temenggor, where it was the most commonly photographed mammal (Lynam et al. 2007). Sambar was found in only one of 14 sites (seven virgin jungle reserves and seven adjacent logged areas) in a study assessing the importance of virgin jungle reserves, making it the most localized mammal species detected. The occupied site was a logged area (Laidlaw 2000). None was found in the Jerangau Forest Reserve, from which it was suspected to have been eradicated by hunting (Azlan 2006). Sambar is still regularly camera-trapped in Johor (M. Gumal pers. comm. 2008) but seems to be generally rare throughout other areas. In Indonesia, there are various recent records from Sumatra, e.g. Bukit Barisan Selatan National Park (O'Brien et al. 2003), Gunung Leuser (van Schaik and Griffiths (1996), and Way Kambas National Park (S. Hedges pers. comm. 2008), but in these areas it was recorded rather rarely by comparison with other ungulates. Sambar was nearly six times as abundant in areas of Bukit Barisan Selatan National Park with low than with high human population density within 10 km of the park boundary, suggesting low resilience to human presence, presumably the effects of hunting, and reduced populations overall (O'Brien et al. 2003). In Way Kambas National Park the population is reportedly stable (G. Semiadi pers. comm. 2008). Overall in Indonesia (Sumatra and Borneo), Sambar is not under major threat except where forest is being converted (i.e. most of the lowlands). Reflecting this, E. Pollard (pers. comm. 2008) considered the status of Sambar, in suitable habitat, to be much more healthy there than in Cambodia and Viet Nam, although there is a significant market for Sambar antlers in Sumatra (D. Martyr pers. comm. 2008). On Borneo, Sambar has been extirpated or persists only at extremely low densities in Lambir Hills National Park, Sarawak, Malaysia (Azlan and Engkamat Lading 2006). An intensive survey of Batang Ai National Park, Sarawak, found Sambar only once; this was stated to be a genuine reflection of actual rarity, and to result from intensive hunting at salt-licks (Meredith 1995). Across Sarawak it is seriously reduced by hunting (Bennett et al. 2000, see Threats; A.C. Sebastian pers. comm. 2008). Some better populations survive in Sabah, Malaysia, e.g. Danum Valley (Heydon 1994), Tabin Wildlife Reserve (Matsubayashi and Sukor 2005), and Deramakot Forest Reserve (Matsubayashi et al. 2007) but populations are in general also much reduced there (Bennett et al. 2000). In the Indonesian part of Borneo populations seem to be less depleted, particularly in East Kalimantan, which is believed to hold the largest Sambar populations among all Kalimantan provinces; in the Penajam Paser Utara district of East Kalimantan, approximately 60 hunted animals are sold monthly in several traditional markets (Semiadi et al. 2004; G. Semiadi pers. comm. 2008).
Population Trend: Decreasing

 Habitat and Ecology:

No large Indian ungulate has adapted itself to a wider variety of forest types and environmental conditions than has Sambar (Schaller 1967). Within India, Sambar occurs in the thorn and arid forests of Gujarat and Rajasthan, in the moist and dry deciduous forests throughout peninsular India, in the pine and oak forests at the Himalayan foothills, and in the evergreen and semi-evergreen forests of northeastern India and the Western Ghats (Sankar and Acharya 2004; N.S. Kumar pers. comm. 2008). Outside India it extends into temperate-latitude and alpine-zone woodlands of Taiwan (Lin, C.-Y. and Lee, L.-L. pers. comm. 2008). This habitat flexibility is permitted by its broad diet: Sambar has been documented to eat 130–180 species of plants in India alone (Schaller 1967; Johnsingh and Sankar 1991; N.S. Kumar pers. comm. 2008), with food requirements less specialised than those of other deer (Schaller 1967). Sambar grazes or browses depending upon the forage available at that time (Schaller 1967; Richardson 1972; Martin 1977; Bentley 1978; Dinerstein 1979; Kelton and Skipworth 1987; Ngampongsai 1987; Sankar 1994; Semiadi et al. 1995). Across the very wide altitudinal range occupied in Taiwan, the diet varies reflecting major changes in plant communities (Lin, C.-Y. and Lee, L.-L. pers. comm. 2008). Sambar was found to live in much higher densities in moist than in dry deciduous forests of Nagarahole National Park (Karanth and Sunquist 1992) and there are no doubt also patterns of habitat selection across the rest of its range. Although the highest densities of Sambar so far recorded were in the semi-arid forests of Ranthambore (Kumar 2000), across most of its Indian range Sambar seems to thrive best in well-watered, moist deciduous hilly terrain (N.S. Kumar pers. comm. 2008). The Sambar occurs up to at least 3,825 m asl on Siouguluan Mountain, the highest peak of the Central Mountains in Taiwan; elsewhere on the island it ranges down to 150 m asl, mostly living at 2,000–3,500 m (Lin, C.-Y. and Lee, L.-L. pers. comm. 2008). It occurs up to 3,000 m asl on Gunung Kinabalu, Sabah, Borneo (Payne et al. 1985). In Myanmar, recent camera-trap photographs spanned the range of 0–2,150 m asl (Saw Htun pers. comm. 2008). Sambar is largely restricted to hilly terrain in the Terai Arc Landscape (Johnsingh et al. 2004), although how much this reflects real habitat selection and how much it is an artificial pattern produced by human effects (habitat conversion and hunting) is unclear. More widely in India, there does seem to be a marked preference for undulating terrain (N.S. Kumar pers. comm. 2008). Kushwaha et al. (2004) found that in Kumaon Himalaya (India), Sambar usage was greater of the higher than the lower altitude area. However, it makes wide use of plains areas elsewhere, where these have not been destroyed, e.g. the Hukaung Valley in Myanmar (J.W. Duckworth pers. comm. 2008).In Borneo, while Payne et al. (1985) considered Sambar “most common in secondary forests of gently-sloping terrain” they also knew of occurrence in “tall dipterocarp forests on steep terrain and in swamp forests”. In Thung Yai, Thailand, Sambar signs were twice as abundant in lowland forest as in montane forest, although this difference was not statistically significant (Steinmetz et al. 2008). In Southeast Asian regions of dense evergreen closed-canopy forest, Sambar is highly tolerant of forest degradation: indeed, much higher numbers are found in encroached stands than in pristine forests, if hunting is under control (Rijksen 1978; Heydon 1994; Stuebing 1995; Davies et al. 2001). In Sabah, Sambar was camera-trapped in both mature and young forest stands (Matsubayashi and Sukor 2005). All the remaining Lao populations are centred around areas with extensive open, or at least broken, habitat amid forest (Duckworth et al. 1999), and the general paucity of records from the interiors of large blocks of closed-canopy evergreen and semi-evergreen forests, which generally support the least depleted large mammal populations, probably reflects natural patterns of habitat occupation (R.J. Timmins pers. comm. 2008). Similarly, in extensive tracts of deciduous dipterocarp forest the species occurs primarily around patches of denser habitats, and this seems to be a natural rather than hunting-led pattern (Timmins and Ou 2001). In this sense, in Indochina Northern Red Muntjac Muntiacus vaginalis is much more tolerant and versatile than is Sambar. In southern and central India where much of the forest is deciduous, Sambar is a true forest ungulate and conspicuously avoids disturbed and open forests, and is highly sensitive to any sort of forest resource extractive activities (N.S. Kumar pers. comm. 2008). Kushwaha et al. (2004) found that in Kumaon Himalaya (India), Sambar was primarily in areas of high tree and herb density with low shrub density. O'Brien et al. (2003) found no statistically significant difference in Sambar abundance between areas within 1 km inside the boundary of Bukit Barisan Selatan National Park, Sumatra, and those deep in the interior of the park. Caution is needed in inferring that there is truly no association of Sumatran Sambar with forest edges (which would be astonishing by comparison with its habitat use elsewhere) because the study does not specify whether this ‘non-significant’ result might simply reflect methodological factors rather than a genuine lack of biological effect (see, e.g., Johnson 1999). Despite the population rises that occur in post-logging forest, there is no evidence that Sambar can survive landscape-level conversion to exotic plantations or other non-forest land-uses, although many live in coconut plantations (G. Semiadi pers. comm. 2008) and it is relatively common in immature Acacia mangium plantations within the matrix of plantation and natural forest of the Sarawak Planted Forests area, Bintulu (Belden Giman pers. comm. 2008). A study of coffee estates around Bhadra Wildlife Sanctuary, India, recorded Sambar only in those coffee areas within 1 km of the sanctuary’s boundary (Bali et al. 2007). In forest protected areas of Thailand such as Khao Yai National Park and Phu Khieo Wildlife Sanctuary, Sambar populations are often concentrated around anthropogenic grass and scrub (e.g. sites of former villages) rather than in the forest itself (Trisurat et al. 1996; Lynam et al. 2001) and this also seems to be true on remote parts of the Bolaven Plateau of south Lao PDR (Evans et al. 2000). In the Annamite mountains of Lao PDR and Viet Nam, Sambar seems often to be associated with degraded valley bottom areas, largely the result of long human influence: although the effects of a permanent water source and differences in terrain in determining distribution are difficult to disentangle (R.J. Timmins pers. comm. 2008). Sambar regularly visits salt licks (e.g. Matsubayashi et al. 2007), perhaps especially when growing new antlers. This predictability exposes it to high levels of hunting, where this is not effectively controlled. It seems that within an area, not all sort of licks are visited. At the Seima Biodivesity Conservation Area, Cambodia, Sambar photographs were initially very few at mineral lick camera-traps. But in 2007, cameras placed on trails and at some permanent water sources (especially in semi-evergreen forest) recorded Sambar much more often. Checking tracks at more than 40 licks indicated that some were used by Sambar and Red Muntjac, but not Gaur Bos gaurus or Asian Elephant Elephas maximus, while the big ones with many cattle and elephant prints had few Sambar prints (E. Pollard pers. comm. 2008). Sambar was considered to be mostly nocturnal by Kawanishi and Sunquist (2004), to show ‘no pattern’ (meaning unclear; perhaps intended to imply no significant variation through the 24-hour cycle, although this is itself a pattern) by O’Brien et al. (2003), and to be cathemeral by van Schaik and Griffiths (1996). In fact, its activity pattern may vary across sites, but in general it is mostly crepuscular, with significant nocturnal activity as well (Schaller 1967; J.W. Duckworth pers. comm. 2008). Sambar is essentially non-social, stable groups being at most family associations (Schaller 1967; Karanth and Sunquist 1992). However, it is often in groups, presumably temporary, of 12–30 in dry deciduous and semi-arid forests of India (e.g. Bandipur, Pench, Melghat and Ranthambore) (N.S. Kumar pers. comm. 2008), and the same is true around the secondary grasslands within Khao Yai National Park, Thailand (J.W. Duckworth pers. comm. 2008). During peak summer, such associations of up to 80–100 near large waterbodies are not uncommon in Pench in central India (N.S. Kumar pers. comm. 2008). A detailed review of information on grouping in Sambar is in Sankar and Acharya (2004). Results from many areas show a strong bias in sex ratio to females, perhaps reflecting selective predation; alternatively, stags may be more vulnerable to stress (Sankar and Acharya 2004). Breeding is rather seasonal in most areas where studied, for example Schaller (1967) reported that in Kanha the rut spreads over a period of at least seven months with a peak in November–December. Stags during the rut sometimes can cover large distances, up to 10–20 km in one night. Further information for India is reviewed in Sankar and Acharya (2004). The bucks of Formosan Sambar have a seasonal antler cycle, and the mating season is from June to January with a peak in September–October. Males wallow and mark with scent glands in this season. Fawns are born (usually singly) from March to August, with a peak in May and June (Lin, C.-Y. and Lee, L.-L. pers. comm. 2008). Predation appears to be the major cause of Sambar mortality. In Bandipur, Sambar is one of the most important prey in terms of the biomass taken by Tiger Panthera tigris (31%), Leopard P. pardus (8%) and Dhole C. alpinus (13%) (Andheria et al. 2007). In Bandipur Sambar remains were found in about 22.3% of Tiger faeces, 6.1% of Leopard faeces and 7.3% of Dhole faeces (Andheria et al. 2007). In adjoining Nagarahole, proportion of Sambar in Tiger’s diet was about 25–29%, derived from faeces and kill data respectively (Karanth and Sunquist 1995). In Tadoba-Andhari, Sambar together with Gaur contributed to nearly 70% of the prey biomass consumed by Tiger, whereas in Pench-Maharashtra it was 80% (Karanth and Kumar 2005). Sambar occurred in 50% of Tiger scats in Pench-Maharashtra (Karanth and Kumar 2005). In well protected forest reserves where different body-sized ungulates are abundantly available, Sambar together with Gaur are selectively predated by Tiger.
Systems: Terrestrial

 Major Threat(s):

 Habitat encroachment and hunting are both widespread in the Sundaic region and in much of the rest of Sambar’s range. Indeed, these two threats are often associated and all Borneo’s ungulates suffer from the increased hunting that often accompanies logging (Bennett and Dahaban 1995; Bennett and Gumal 2001), such that hunting to local extinction is the chief detrimental effect of logging on ungulates in Sarawak (Bennett and Gumal 2001). This conclusion probably is applicable across the range of Sambar, within which it is heavily hunted (e.g. Duckworth et al. 1999; Walston 2001; Tungittiplakorn and Dearden 2002; Linkie et al. 2003; Santiapillai and Wijeyamohan 2003; Tilson et al. 2004; Rao et al. 2005; Steinmetz et al. 2008; Saw Htun pers. comm. 2008). It is among the most sought-after wild meats in Indonesia (Semiadi 2005), Sarawak (Belden Giman pers. comm. 2008), and Viet Nam (B. Long pers. comm. 2008), and during 1988–1993 it was a common and preferred food in rural Lao PDR (Duckworth et al. 1999, Table 1), and is probably near the top of chosen wild meats throughout most of its range (GMA Indonesia workshop); however, it is less favoured in Taiwan compared with Formosan Serow Capricornis crispus and Reeves’s Muntjac Muntiacus reevesi (Chang Shih-Wei pers. comm. 2008). However, it is now so rare in Nam Ha National Protected Area, Lao PDR, that a more recent hunting study found it to be eaten only rarely (Johnson et al. 2003). In Vietnam, muntjac meat is now often served in wildlife restaurants as Sambar, because real Sambar meat is now so difficult to procure (B. Long pers. comm. 2008). These declines in consumption are probably representative of all Lao PDR and Viet Nam, and of increasing proportions of the rest its range, reflecting population losses (see Populations). There are major, ongoing, declines in at least Vietnam, Lao PDR, Cambodia, Thailand, Malaysia, Myanmar, Bangladesh and Nepal, and probably Sri Lanka and Indonesia (see Populations) which can plausibly only be driven by hunting, because suitable habitat for Sambar is abundant in these countries but is almost or actually bereft of the species. Even in India, poaching has seriously depleted the abundance of large mammals in most areas; even in high-profile areas such a Corbett Tiger Reserve and Rajaji National Park there are still some instances of poaching (Johnsingh et al. 2004). An intensive study of hunting on mammals in two Hmong villages of northern Thailand classified quarry species into tiers representing the sequence of loss through over hunting; Sambar fell in the second tier (aside, e.g., big cats Panthera) in the extinction sequence (Tungittiplakorn and Dearden 2002). This study found that Sambar persisted anomalously long in one place under an “exceptional village wildlife hunting regulation promulgated by the village headman until 1994”, but was generally extinct in the landscape. A study at Nagarahole National Park compared an area which was only moderately hunted with a heavily hunted site, but found no significant difference in Sambar densities between the two areas (Madhusudan and Karanth 2002). This was speculated to reflect the greater difficulty of hunting Sambar (highly dispersed and usually in rather thick vegetation), despite it being a prized species, than, particularly, Chital Axis axis, a herding species often out in the open, and the benefit to poachers of getting in, making a kill, and getting back out again as fast as possible to reduce detection chances; therefore, this pattern seems to have arisen through hunter choice. In the context of Indochina, even the heavily-hunted site would, however, rank as lightly hunted (J.W. Duckworth pers. comm. 2008, based on visit and discussion). For these two reasons, their finding, of an apparent high resilience of Sambar to hunting, is not applicable to the rest of the species' range (particularly outside well-secured protected areas). However, somewhat reflecting this, Sambar’s status in non-Sundaic South-east Asia (still widespread but in ongoing steep decline) relative to other large deer ecologically more similar to Chital (herding; often outside dense forest), Eld’s Deer Rucervus eldii and Hog Deer Axis porcinus (extinct across most of their former range through former steep declines) does indeed suggest a broad applicability of a conclusion that Sambar is somewhat less rapidly reduced by hunting than are other sympatric deer, excepting muntjacs Munitacus spp. Deer are hunted in India using snares, dogs, and guns (Jathanna et al. 2003; Kumara and Singh 2004) and these methods are general across their range. Some hunting is for village consumption of meat, but most (at least in South-east Asia) is probably sold commercially, to the affluent urban classes (e.g. Duckworth et al. 1999; Walston 2001; Kumara and Singh 2004), and to itinerant labourers, such as logging crews in at least Sarawak (Bennett and Gunal 2001) and gold-diggers and rattan-cutters in Myanmar (J.W. Duckworth pers. comm. 2008). Such commercially-driven hunting causes major declines (e.g. Steinmetz et al. 2008) because the market is limitless. Adult males suffer additionally because antlers are widely displayed as trophies and are used in traditional medicine (e.g. Martin 1992, Baird 1995). At least in Lao PDR, their market value means that, compared with muntjacs, a high proportion is sent to towns and traded internationally, and many are sold openly in tourist centres such as Louangphabang (Duckworth et al. 1999). In Taiwan, in the past live males were sought for farming velvet; this was far more valuable than the meat. Presently, velvet can be bought legally and cheaply from farms and hunting for velvet is now insignificant, as is, here, hunting for internal organs (Chang Shih-Wei pers. comm. 2008). Hunting pressures and patterns vary across Sambar’s range: three areas (Indochina; Malaysian Borneo; Taiwan) are profiled below, and all are very different not just from each other, but from southern and central India, where (excepting Taiwan) the species has by far the healthiest conservation status. But even across India, hunting for Sambar meat is almost ubiquitous, even occurring within well-secured protected areas, and to greater extents in other protected areas and outside the protected area system (N.S. Kumar pers. comm. 2008). Vietnam, Lao PDR and to a lesser extent Cambodia and northern Thailand are apparently the areas of Sambar’s range where generalised mammal hunting is heaviest. Sambar is hunted within a thriving hunting and wildlife trading culture in this area and in adjacent countries such as China and Thailand, involving many land vertebrate species, along with other forest products such as orchids and Aquilaria resin (e.g. Compton and Le Hai Quang 1998; Compton et al. 1999; Noreen and Claridge 2001; Smith and Xie Yan in press). Such hunting reaches all areas, although in large rugged mountain forest blocks trade-driven hunting of species valued only for their meat (rather than high-value, low-weight, generally medicinal products that can be carried out efficiently from even the remotest areas) is still limited by economics of accessibility. However, the ability of these areas to protect Sambar at the regional scale is questionable, because no areas of rugged evergreen forest are known with high Sambar densities: such habitat is rather marginal. The regional wild meat trade has little likelihood of abating as long as there are any of the most resilient species (pigs, muntjacs, and civets) to be hunted. The human population of Vietnam is more than 84 million, that of China more than 1,000 million. Together, they comprise an enormous market for wildlife products. For example, tens of millions of wild turtles are imported, legally and illegally, into China annually (van Dijk et al. 2000 and papers therein). There has been no comparable study of ungulate trade levels. Within the Northern and Central Annamites, every square kilometre of Viet Namese forest and of Lao forest that is within 5 km of the Viet Namese border probably has snares capable of capturing Sambar set in it every year (Timmins et al. 2007; R.J. Timmins pers. comm. 2008). Intensity in some areas probably reaches many thousands of snare-nights per km˛ per year (Timmins et al. 2007). Snaring is less intensive in Lao PDR at least away from the Viet Namese border, but is increasing dramatically (Timmins and Robichaud 2005; W.G. Robichaud pers. comm. 2007; R.J. Timmins pers. comm. 2008). Hunting intensity has in recent decades been generally lower in southern Viet Nam than in Lao PDR and northern Vietnam, leaving many wildlife populations less depleted (e.g. Le Xuan Canh et al. 1997; Timmins and Duckworth 2000; Polet and Ling 2004 ), but the general increase in regional hunting intensity and economic wealth indicate that hunting intensity in this region must also be increasing. Rapid economic development and expanding wealth, particularly South-east Asia and China, are increasing the demand for wildlife meat and ‘medicines’ (e.g. Timmins et al. 2007). Timmins et al. (2007) highlighted a common misconception, especially in documentation of development projects in the region, that poverty is a principal cause of biodiversity loss: as they pointed out for Saola Pseudoryx nghetinhensis, the main driver of threats to wild ungulates in Indochina, at least for the mid-term, is not rural poverty but increasing urban wealth in Indochina and China. In Viet Nam “the free market economy has resulted in feverish periods of trade in wild species nationwide, with negative impacts on biodiversity” (Government of the Socialist Republic of Viet Nam 2004). In Cambodia, the same factor has fostered a thriving bush meat market and hunting of species for international wildlife trade (Timmins and Ou 2001; Timmins 2006; Maxwell et al. 2007), and the intensity of hunting there for some species (Sambar included) is likely to exceed even that in Viet Nam. This reflects the logistical ease of hunting and trading with few controls and the relative abundance of high-value quarry species (R.J. Timmins pers. comm. 2008). There is a rapid, ongoing, expansion of wealthy social strata in Lao PDR, Cambodia, and Viet Nam, directly accelerated by illegal trade in timber, wild meat, and other forest resources, and by the economics of large infrastructure projects (e.g. Nan Theun 2 hydro-electric power project), creating a significantly greater in-country demand for luxury meats such as venison (W.G. Robichaud pers. comm. 2007). Markets along major roads such as route 13 through Lao PDR (e.g. at Ban Namthon) have expanded, not contracted, in the last 15 years and have a huge (albeit unquantified) turnover of wild meat; at this specific market, multiple stalls are selling dried deer meat daily. The effects of hunting in Lao PDR, Viet Nam, Cambodia and Thailand have been exacerbated during the last two decades by habitat loss and various socio-economic factors, of which the most biologically significant are discussed in the Red List account for Large-antlered Muntjac M. vuquangensis. The most significant constraint to Indochina’s Sambar populations is the long-term uncertainty of success in protected areas, even in those currently effective in conservation management. Even the most successful protected areas face an uncertain future with the possibility of degazettement of conservation status of parts of them, the lack of, or possible future loss of, adequate external funding necessary to maintain high standards of management, the lack of, or possible future loss of, political support necessary to uphold high protection standards and the uncertainties of maintaining a motivated and well-trained staff. On Borneo, Sambar are also widely and heavily hunted, largely for meat consumed in-country (Bennett et al. 2000; Mohd Azlan J. pers. comm. 2008; Belden Giman pers. comm. 2008; G.M. Fredriksson pers. comm. 2008; A.C. Sebastian pers. comm. 2008; Siew Te Wong pers. comm. 2008). Traditionally, hunters used traps, dogs and spears, and blowpipes. Hunters in remote parts of the interior still use these methods, but most animals now die by gunfire. Sarawak held 60,000 legally registered shotguns while Sabah held almost 13,000 in the mid 1990s (Bennett et al. 2000). Methods used in Indonesia include guns and spotlights along logging roads, snaring, and dogs (G.M. Fredriksson pers. comm. 2008; S. Hedges pers. comm. 2008). Snaring is still very heavy outside well-secured areas, but shotguns remain the even bigger problem (Bennett et al. 2000; Mohd Azlan J. pers. comm. 2008; A.C. Sebastian pers. comm. 2008; Siew Te Wong pers. comm. 2008). Over most of (at least) Malaysian Borneo, “in general, everyone will hunt and eat anything” (Bennett et al. 2000). Much hunting is for the market rather than local use, so there is a limitless demand. Bennett et al. (2000) and Bennett and Gumal (2001) profiled the hunting of ungulates in Sarawak in the mid 1990s thus: Sambar (along with Bearded Hog Sus barbatus and muntjacs Muntiacus) is heavily shot for wild meat trade across Sarawak, and these ungulates are the species most sought by the restaurants; Sambar in particular is an important source of meat for logging company employees in Sarawak. Wild meat is widely sold in towns, villages, restaurants and logging camps throughout Sarawak; 250 sales outlets were estimated across the state in 1996, with an annual trade worth 3.75 million USD. Along the Rejang river alone, in the mid 1980s, 1,500 Sambars were sold per year (Caldecott 1988). The market for meat is great and probably expanding. Bennett et al. (2000) considered the effects of this onslaught upon Sambar through a lengthy hunting study in February 1993 to June 1995, in both Bornean states of Malaysia (Sarawak and Sabah), each with eight study areas. Sambar constituted only 6.7% of animals killed in Sabah, but 35% by weight of wild meat; negligible numbers were taken in Sarawak, reflecting the major declines that had already reduced the species to rarity. Index counts of signs show an inverse relationship between sign index and hunting pressure so strong that the latter over-rode the effects of habitat variables in determining their densities. Hunting was so ubiquitous, even in protected areas, that in the two states, only one site (part of Danum Valley, Sabah) could be found as a control, where there was primary forest but negligible hunting levels. They concluded that “the only single factor offering any effective protection for [quarry species of] wildlife is difficulty of access”; whether animals are protected or non-protected species, inside or outside protected areas, had no significant restraint on hunting levels. There were, then, no official controls on commercial hunting of non-protected species. The Master Plan for Wildlife in Sarawak (Wildlife Conservation Society and Sarawak Forest Department 1996) introduced a strict ban on all wildlife trade, shotgun ownership and cartridge sales in the late 1990s (Bennett and Gumal 2001). Substantial numbers of Sambars are, nevertheless, still killed in the state, as they are elsewhere in Borneo. Moreover, in the interim, much forest has been lost and Sambar populations in the smaller and more isolated areas that remain must now be even more vulnerable to local extinctions. Taiwan contrasts with the rest of Sambar range in that hunting has been so successfully reduced that populations are now increasing in large parts of the island. It was formerly a big problem, for venison and, for medicinal purposes, velvet and penis. The reduction during and since the 1990s reflects five factors. Hunting of Sambar was banned in 1989 through the Wildlife Conservation Act. Velvet prices fell steeply during 1985–1990 with deer TB in 1985, reducing returns from poaching absolutely an effect exacerbated by rapid economic development in 1980s leading to higher daily wages. Former logging road systems are deteriorating within protected areas, making removal of deer meat less attractive. The general awareness of ecological conservation has much enhanced. And eco-tourism through community forestry plans has given alternative employment to former hunters (all information: Lin, C.-Y. and Lee, L.-L. pers. comm. 2008). Other sympatric ungulates are hunted broadly as avidly as Sambar, but there is suggestive evidence from two areas of Thailand (Kuiburi National Park and a 20 km˛ sector of Thung Yai Naresuan Wildlife Sanctuary) that when Sambar densities are very much reduced, recovery may be more difficult than with Gaur Bos gaurus and muntjacs. In both these areas the institution of effective protection fuelled a rise in sign density, and presumably animal populations, for Gaur and muntjacs, but not for Sambar (Steinmetz et al. 2007, in prep.; R. Steinmetz pers. comm. 2008). If this finding is widely applicable (and B. Long pers. comm. 2008 suspects a similar contrast between Sambar and other ungulates in a protected part of Cambodia), the implications for Sambar populations in Lao PDR, Viet Nam, northern Thailand, Sarawak and other areas where populations are mostly down to scattered tiny numbers, may be profound.
Over many centuries, the loss of wilderness to human settlement, cultivation and industry has resulted in massive loss of potential Sambar range, particularly in the densely-settled countries of India, China, Bangladesh, Thailand, and Viet Nam. Throughout the range of Sambar, natural and semi-natural habitats (forest, scrubland and grassland) continue to be converted to anthropogenic habitats, degraded (e.g. stands are being changed in character through logging) and fragmented (broken into smaller blocks). The relative proportions of these activities vary across the range as does the percentage of former forest land so affected. The role of these habitat changes as a direct stimulant of the major declines in progress is unclear. Sambar is highly tolerant of forest degradation: indeed, much higher numbers are found in encroached stands than in pristine forests, if hunting is under control (Rijksen 1978; Heydon 1994; Stuebing 1995; Davies et al. 2001). Degradation per se is unlikely to be a significant threat. Many Indian forest areas are severely encroached by exotics such as Lantana camara, Parthenium spp. and Chromolaena odorata, and these are suspected to effect major changes to forest structure (Hiremath and Sundaram 2005): but their effects on Sambar populations warrant further study. There is no evidence that Sambar can survive complete conversion of landscapes to non-forest uses such as rice paddies, and its ability to use tree plantations in the absence of natural forest is unclear. Probably, like most deer, provided strips of semi-natural vegetation survive (e.g. along streams and on steep slopes) are present, it can use them to some degree. In most of its range (parts of India excepted) it is unlikely that habitat extent is the determining factor of current Sambar numbers, because large tracts of apparently suitable habitat now support no or only tiny numbers of Sambar. By far the most serious effect of current habitat trends is that fragmentation opens up wilderness areas to people, making it easier to hunt throughout them and cheaper to remove bulk wildlife products such as Sambar meat. In both Malaysian Borneo and Lao PDR the ‘natural protection’ afforded to an area through difficulty of access has been considered to be the chief factor allowing populations of large, hunted, mammals to survive (Timmins and Duckworth 1999; Bennett et al. 2000), and this is probably widely true for Sambar outside India, Nepal and the few other areas where active protective measures are effective. In sum, the proximate cause of the major ongoing Sambar declines in most of its range is hunting, for which habitat conversion and fragmentation exacerbate the effects; and these processes are resulting in an ever-smaller maximum possible population, were the hunting issues to be successfully addressed. Threats other than hunting and, locally, habitat loss, seem to be insignificant at the species level. Hybridization with Javan Rusa Rusa timorensis has been documented (Van Mourik and Schurig 1985), and is apparently a localized threat, where the latter has been introduced into Sambar range, in Pewajam District, East Kalimantan (G. Semiadi pers. comm. 2006). Hybridization has also been suspected in Sumatra, with the introduced Chital Axis axis (G. Semiadi pers. comm. 2006). However, Sambar and Chital co-occur extensively in India without hybridizing. Sambar is an occasional crop pest in mainland China, notably around the Ailoshan range (J. Fellowes pers. comm. 2008), India (N.S. Kumar pers. comm. 2008), Lao PDR (Duckworth et al. 1999), Myanmar (J.W. Duckworth pers. comm. 2008), Sumatra (D. Martyr pers. comm. 2008), and probably throughout its range. There is no evidence that Sambar is a serious crop pest away from the fields directly abutting forests. Nonetheless, crop damage is often stated by affected farmers to lead to a requirement for lethal control. However, the ready market for meat means that this is most probably an excuse for killing that would enthusiastically be undertaken anyway (N.S. Kumar and J.W. Duckworth pers. comm. 2008): it is therefore unlikely that control of crop-degrading animals is a significant cause of the species' decline anywhere in Sambar’s range. Grazing competition with domestic stock seems to be much less of a threat to Sambar than to sympatric Gaur and Chital: in fact the key study quantifying just how serious a threat such competition is in Indian protected areas (Madhusudan 2004) found that it had negligible effects on Sambar densities, presumably because Sambar is an efficient browser. Threats to southern India’s forest ungulates by competition with domestic stock grazing within protected areas are exacerbated where dung is collected for export to adjacent coffee areas. Fuel wood removal may also be at levels sufficient to disrupt nutrient cycles of the habitat (Madhusudan 2005). In the Indian Terai Arc Landscape, lopping of tree branches for cattle fodder is a threat to Sambar as it effectively removes most browsable vegetation from the landscape (B. Long pers. comm. 2008). These are probably more serious concerns for Sambar than is grazing by domestic stock.

 Conservation Actions:

 The Sambar is found in many protected areas throughout its range, although in most of these areas this legal status has not stemmed declines and local extinctions from hunting. Similarly, although it is protected from hunting by legislation, even outside protected areas, in most or all range states, these laws are challenging to enforce, given the trade demand for meat and antlers (GMA Indonesia Workshop). Currently the law has a strong protective effect in Taiwan (Lin, C.-Y. and Lee, L.-L. pers. comm. 2008) and in various protected areas of India (N.S. Kumar pers. comm. 2008). In South-east Asia, Sambar declines have lagged behind those of other sympatric deer (excepting muntjacs). There is no doubt that within a decade or two Sambar will be, like Hog Deer and Eld’s Deer are already, almost absent from South-east Asia unless effective protection from trade-driven hunting is instituted. In India the species’s status is less grim, and this is dependent upon continuation of the current protected areas system, and the expansion of effective threat reduction into a greater proportion of protected areas: in many Indian protected areas, poaching is rampant and in such vulnerable areas Sambar numbers are still declining. Given the major challenges combating hunting of such a desired animal, long-term survival is most realistic in large protected areas (exceeding 1,000 km˛ where possible), lacking people living within them, with an adequate number of motivated and capable staff (Anak Patannavibool pers. comm. 2008; A.J. Lynam pers. comm. 2008). The precise ways for protected areas to function effectively in the conservation of large mammals depend upon their own specific circumstances, but field presence of staff, dispersed across the area, is likely to be important in almost every area. For example, Lynam et al. (2006) found in Khao Yai National Park, Thailand, that encounter rates of tracks and signs of Sambar (and other large mammals) decreased with distance from ranger substations. Deep inside the forest to the south and east, far away from ranger substations, Sambars were locally extinct. Similar patterns are shown by current Sambar distributions in other Thai protected areas such as Khao Sok, Phu Khieo and Kaeng Krachan (A.J. Lynam pers. comm. 2008). In some protected areas (recent examples being Bhadra and Kudremukh in Karnataka), formerly forest-dwelling human communities have resettled closer to markets, clinics, school and other services (Karanth and Karanth 2007). This expands habitat available to Sambar (that which was formerly fields and houses) but more significant is the removal of a source of hunters and the cessation of grazing competition with domestic stock (see Karanth et al. 2006). Such conservation interventions have helped recovery of Sambar populations in this region. Increasing development aspirations mean that out migration is likely from further areas, e.g. Mudumalai. Most of the remaining forest within its Chinese range are already established as nature reserves (M.W.N. Lau pers. comm. 2008). As the majority of the Sambar population is in South Asia, conservation activities are important there. It is also important that populations are maintained at least in several areas in Southeast Asia, to maintain Sambar presence through its historical range, to preserve Sambar genetic diversity and through its importance to the ecosystems within which it lives. Given the regional pattern of threats and current successes to date, implementing effective conservation interventions is considerably more challenging, and therefore a higher priority in the Southeast Asian region. The most important conservation measure for Sambar in Cambodia is the continued strengthening and support of conservation efforts in Siema Biodiversity Conservation Area and the Srepok Wilderness Area, and to a lesser extent support of protected area conservation management in the northern plains and central Cardamoms. Likewise in Vietnam, the most important conservation measure is the continued support of protection efforts within Cat Tien National Park, and in Thailand it is to continue the consolidation of the ‘Western Forest Complex’. In Myanmar the Hukaung Valley is an exceptional area of lowland plains forest, grassland, and wetlands. Part lies within the Wildlife Sanctuary and some of the rest of the outstanding plains habitat is within an enormous proposed extension. Through its size and mostly little-encroached condition, this is the most outstanding remaining landscape-level floodplains habitat for very large mammals remaining in tropical Asia, and although wildlife populations are highly depleted, warrants the strongest effort to conserve it. It faces a number of daunting challenges (J.W. Duckworth pers. comm. 2008). In Lao PDR, any of a large number of areas could become key Sambar conservation sites, but there is as yet no precedent in the country for effective conservation of high-trade-value large mammals. Xe Pian, Nam Et–Phou Louey, and Nakai–Nam Theun NPAs could all be highly significant areas for the species, but so could almost any other NPA; the reality is that the most success with Sambar conservation is likely to come through identifying areas with positive underlying situations to achieve conservation, rather than identifying the area with the most intrinsic importance for Sambar as the focus for efforts. Captive breeding as it is being developed in East Kalimantan. A captive herd of 223 heads in 10 ha enclosure was established in 1998 in Penajam District, as a pilot project. Several private herds exist in the province, such as at Bearau and at Nunukan, both with more than 20 heads (G. Semiadi pers. comm. 2008). Captive Sambar herds are also widespread in mainland Asia.